Summary
In this chapter, we have explored what model parameters are and how a sweep job can be leveraged to tune hyperparameters that are defined for a given model. We have also explored options for setting up sweep jobs based on the search space and sampling methodology selected. AMLS provides the ability to sweep across the search space to tune a model, automating the process of hyperparameter tuning on a compute cluster, which will shut itself down in the idle period after the trials are completed, consuming compute resources wisely.
In addition to setting up a sweep job, you have been able to review your results in the Studio as well as in the code – providing intuitive insight into the best-performing model for your use case. Now that you have completed the chapter, be sure to turn off your compute resources to save cost.
In the next chapter, we will show you how to leverage AMLS to take over the time-consuming task of model development. This functionality is not...
